Latest Articles for Training Data

Innovative strategies for safe off-road driving using intelligent control systems.

Research reveals how larger AI language models predict brain responses to language.

Learn how to manage overfitting for better model performance.

Research examines prompting techniques to improve trustworthiness of language model outputs.

A model designed for effective multi-table data querying and structured outputs.

A new self-supervised method enhances logical reasoning without extensive labeled data.

This article explores how training data affects image generation by text-to-image models.

New methods for selecting examples enhance performance in language models.

New graph-based method enhances entity extraction from various document types.

A study on how diverse training data improves text style transfer.

Innovative methods using synthetic data improve anomaly detection in various fields.

Improving language models with knowledge for better medical understanding.

This article discusses the difficulties PINNs face in maintaining mass balance during simulations.

NormMark enhances chatbots by recognizing social norms in conversations.

Discovering surprising behaviors in neural networks during training.

Exploring ASR performance across different accents and its implications.

This article analyzes the portrayal of non-cisgender identities in image generation technologies.

This article examines gender bias in translation models and methods to improve fairness.

A novel method enhances metaphor detection using limited training data.

WELL simplifies bug localization using weakly supervised learning methods.

Evaluating strategies to manage inappropriate outputs from image generation models.

A method for teaching computers to generate and customize game levels and models.

This study explores new methods for improving math problem-solving using language models.

An overview of recent advancements in text style transfer techniques.

A new method combines machine learning with simulations to generate valuable data.

This study focuses on recognizing Arabic dialects using advanced methods and limited data.

Addressing challenges in voice recognition of African names for better user experience.

LOWA enhances object detection by focusing on traits and flexible descriptions.

CoDi enables simultaneous generation of diverse content types from various inputs.

A detailed look at speaker recognition techniques developed for the 2022 competition.

Exploring how ciwGAN can learn and represent phonological features like nasality.

Study assesses privacy risks of generative models in medical data sharing.

Studying how biases in training data influence model behavior and performance.

A new method improves image recognition performance on rare categories.

This research presents a method to balance creativity and factual accuracy in language models.

Polyglot-Ko aims to create high-quality language models for Korean speakers.

This study highlights the need for better recognition of non-binary pronouns in language models.

Integrating user feedback to improve image captioning models for better user experience.

Researchers apply machine learning techniques to understand nanosilicate clusters in space.

A new approach improves text classification across multiple languages.